Victoria’s largest supercomputer powers local search – OpenGov Asia
Magnetic resonance imaging (MRI) technology is a widely used but expensive tool for diagnosing brain injuries and strokes. However, its high costs to supply, install and operate mean that much of the developing world does not have access to it.
Researchers at the University of Hong Kong (HKU) have successfully developed a new magnetic resonance imaging (MRI) technology, 0.055 Tesla Ultra-Low Field (ULF) Brain MRI, which can operate from a standard AC wall outlet and requires no radio frequency or magnetic shielding room. Additionally, a typical conventional MRI machine can cost up to US$3 million, but the ULF-MRI scanner costs only a fraction of that price.
The research team was led by Professor Ed X. Wu, holder of the Chair of Biomedical Engineering and the Lam Woo Chair in Biomedical Engineering in the Department of Electrical and Electronic Engineering, HKU. The research result was published in Nature Communicationand also highlighted in Nature Asia and American scientist.
The HKU team is one of three major academic ULF-MRI research groups in the world, including one based at Harvard/MGH, dedicated to the development of novel ULF-MRI technology. Their goal, shared by researchers like Professor Wu, is to popularize and expand the use of MRI.
As an MRI researcher for over 30 years, Professor Wu is thrilled and derives a strong sense of accomplishment from developing what he calls a “scaled down” MRI scanner that is much more affordable than what is offered in hospitals. The human body is mostly made up of water molecules, which MRI feeds on, said Professor Wu. “MRI is a gift from nature and we need to use it more. Currently, it is underused as a diagnostic tool.
It is estimated that currently more than 90% of MRI scanners are located in high-income countries and that two-thirds of the world’s population do not have access to them. The total number of clinical scanners is estimated to be around 50,000 only worldwide.
The HKU team has made open source ULF 0.055 Tesla brain MRI knowledge design and algorithms available to anyone who wants to further develop the technology or apply it in various fields. This practically opens the door to advancements in various aspects of healthcare delivery in terms of MRI applications. This will be a big area, said Professor Wu, the team has demonstrated the concept and shown the feasibility of a simplified version of MRI. There are many ways to move forward.
Through the use of a deep learning algorithm, the team removed the constraint of conventional MRI, namely the need to be shielded from the outside radio frequency signal, resulting in a bulky setup and not mobile. Existing MRI scanners are essentially giant magnets and need a specially designed room to shield them from outside signals and to contain the strong magnetic fields generated by their superconducting magnets, which require expensive liquid helium cooling systems. The team’s new IT and hardware concept made the latest developments possible.
Professor Wu is convinced that a critical mass of researchers could push back the frontiers of knowledge. He noted that the open source approach is the fastest way to spread knowledge. It is hoped that MRI can be used in fields other than radiology, for example in paediatrics, neurosurgery or the emergency room. The team welcomes more people from scientific, clinical and industrial sectors into research for the benefit of health care, he said.
Together with Professor Gilberto Leung of Neurosurgery and other clinicians at Queen Mary Hospital, his team had validated the results of using ULF-MRI by comparing them with images obtained from a machine Standard 3 Tesla MRI. They could identify most of the same pathologies, including the findings of strokes and tumors, despite the lack of clarity and resolution required for precision diagnoses.
Professor Wu said, “I believe computing and big data will be an integral and inevitable part of future MRI technology. Given the inherent nature of MRI, I believe that widely deployed MRI technologies will lead to immense opportunities in the future through data-driven MRI image formation and diagnosis in the field of health. This will lead to inexpensive, efficient and smarter clinical MRI applications that will ultimately benefit more patients.